# Why some things are darker when wet

> An annotated edition of this paper on Fermat's Library, where scientists and readers explain its ideas, equations and history in the margins. Annotated edition: https://fermatslibrary.com/s/why-some-things-are-darker-when-wet

- **Authors:** John Lekner and Michael C. Dorf
- **Published:** 1988
- **Topic:** Physics
- **Original:** https://s3-us-west-1.amazonaws.com/fermatslibrary/uploads%2F1532418437529-j6l4ebutx4-97c4d8fa5be9ecfa8d630114682e402e%2Fdarkerwhenwet.pdf
- **Pages:** 3
- **Annotations:** 10 comments in 8 threads
- **Featured:** Fermat's Library Journal Club
- **Annotated edition:** https://fermatslibrary.com/s/why-some-things-are-darker-when-wet

## Annotations

### Annotation 1 (page 1)

Beside the passage: “Angstrom has proposed that rough absorbing materials are darker when wet because their diffuse reflection makes possible total internal reflection in the water film covering them, increasing the likelihood of the absorption of light by the surface. His model is extended here in two ways: the probability of internal …”

**Annotation** by Luís Batalha (2018-10-29):

> The swedish physicist Anders Angstrom was the first person to study this effect. He was intrigued by the darkening of soil when in contact with water and published the first proposed explanation of the effect in 1925.
>
> ![](https://upload.wikimedia.org/wikipedia/commons/thumb/9/9a/Anders_%C3%85ngstr%C3%B6m_painting.jpg/225px-Anders_%C3%85ngstr%C3%B6m_painting.jpg)

**Reply** by Mandeep Gill (2018-10-30):

> Huh -- didn't know Angstrom was so recent, seems like the unit would have been named long ago.

**Reply** by Lukas Rødland (2018-11-03):

> It was the grandson of the Ångström that the unit is named after.

### Annotation 2 (page 1)

Beside the passage: “(1 - R)a 1 - p(l-a) (1)”

**Annotation** by Luís Batalha (2018-10-29):

> Note that if you isolate the term $a(1-R_l)$ we get
> $$
> \frac{A}{a(1-R_l)} = 1+(1-a)p+(1-a)^2p^2+(1-a)^3p^3+ \ldots
> $$
> which is the same as the geometric series $\sum^{\infty}_{n=0} x^n = \frac{1}{1-x}, |x|<1$ with $x = p(1-a)$. Thus
>
> $$
> A = \frac{(1-R_l)a}{1-p(1-a)}
> $$

### Annotation 3 (page 1)

Beside the passage: “II. Diffuse Scattering and Internal Reflection An optically smooth surface reflects specularly, whether or not the surface is covered by a film of liquid. A ray of light reflecting at a smooth surface will not be totally reflected at the liquid-air interface. If the surface is rough, its diffuse reflection sends some …”

**Annotation** by Luís Batalha (2018-10-29):

> In the following image we can see the difference between *Diffuse reflection*, where light is reflected from a surface such that a ray incident on the surface is scattered at many angles rather than at just one angle as in the case of *specular reflection*. For ideal diffuse reflecting surfaces the intensity of emitted light (luminance) is the same regardless of the direction we look in the upper half plane.
>
>
>
> ![](https://i.imgur.com/OGGZciA.png)
>
> All the light rays diffused in the blue regions are going to be totally reflected as they hit the liquid-air transition making it impossible to escape to the air.
>
> ![](https://i.imgur.com/EPp839S.png)
>
> Note: If you don't know what total internal reflection is you can quickly learn [here](https://courses.lumenlearning.com/physics/chapter/25-4-total-internal-reflection/).

### Annotation 4 (page 1)

Beside the passage: “Angstrom evaluates p as follows: all the light with an angle of incidence greater than the critical angle , (= arcsinl/n) onto the liquid-air interface will be totally reflected. Thus p may be estimated as the fraction of diffusely reflected light which lies outside the cone generated by rays whose reflection angle is …”

**Annotation** by Luís Batalha (2018-10-29):

> Saying that in a Lambertian Surface the intensity reflected at angle $\theta$ is proportional to $\cos \theta$ (first derived by Johann Heinrich Lambert  in 1760) is equivalent to saying that the surface has the same radiance (power transmitted per unit solid angle per unit projected area) when viewed from any angle.
>
> ![](https://i.imgur.com/hdCzIWP.png)
>
> To prove that, let's analyze the following diagram representing a Lambertian Surface.
>
> The radiance at A is $R_A = \frac{\phi_{source}}{4\pi x_0^2}  $ where $\phi_{source}$ is the luminous flux of the source. At B, the radiance is $R_B= \frac{\phi_{source}}{4\pi x^2}\cos \theta$ since the plane of incidence is tilted at an angle $\theta$ from the normal direction. We just proved that for a degree $\theta$, $R\propto \cos \theta$!

### Annotation 5 (page 1)

Beside the passage: “Angstrom evaluates p as follows: all the light with an angle of incidence greater than the critical angle , (= arcsinl/n) onto the liquid-air interface will be totally reflected. Thus p may be estimated as the fraction of diffusely reflected light which lies outside the cone generated by rays whose reflection angle is …”

**Annotation** by Lars (2020-04-07):

> Can anyone please explain how it subtends such an angle? What I understand from this is something like this: https://prnt.sc/ruroj6. However, I can't understand how it is expressed mathematically.

### Annotation 6 (page 1)

Beside the passage: “P p /2 = Cos 2 OC = 1- 1/n . 2,r I dO sinG cosO”

**Annotation** by Luís Batalha (2018-10-29):

> What we are doing here is calculating the ratio of radiance (power per unit of solid angle per unit of area) that is totally internally reflected (which corresponds in the figure to the blue area)
>
> ![](https://i.imgur.com/WD67C8B.png)
> $$
> p = \frac{R_{blue}}{R_{total}} = \frac{\int^{\pi/2}_{\theta_c}\int^{2\pi}_{0}\frac{\phi_{source}}{4\pi r^2 }\cos \theta d\Omega}{\int^{\pi/2}_{0}\int^{2\pi}_{0}\frac{\phi_{source}}{4\pi r^2 }\cos \theta d\Omega} = \frac{2\pi \int^{\pi/2}_{\theta_c}\cos \theta \sin \theta d\theta}{ 2\pi \int^{\pi/2}_{0}\cos \theta \sin \theta d\theta}
> $$
>
> where $d\Omega = \sin \theta d\theta d\psi$ is the differential solid angle.
>
> The integral is easy to solve since $\frac{d}{d\theta} \cos^2 \theta = -2 \cos \theta \sin \theta$ and so $p = \cos^2 \theta_c$.

### Annotation 7 (page 2)

Beside the passage: “Equation (2) underestimates p: the reflectivities for both polarizations are generally small but not zero for 0 < 0,; see, for example, Ref. 4, Sec. 1.5, or Ref. 5, Figs. 1-3 and 1-4. For a surface which reflects diffusely according to Lambert's law, the probability reflection at the liquid-air interface is J r/2 | dO …”

**Annotation** by Luís Batalha (2018-10-29):

> When light hits a surface we can decompose the propagating electric and magnetic fields in 2 components: the component perpendicular to the surface and the component parallel to the surface. The reflectivities for these 2 components are not the same and in particular, there's a small but non neglectable reflectivity for $\theta < \theta_c$ - which means that there's a percentage of light that is reflected in the liquid-air transition even at angles below the $\theta_c$.
>
> In the following to graphs we can see in blue the reflectivity for the liquid-air transition for the 2 polarizations: perpendicular and parallel. As we can see for both polarizations the reflectivity for $\theta<\theta_c$ is small but not zero, and so we also have to take that into account.
>
> ![](https://i.imgur.com/SjFCGbi.png)
>
> ![]()
>
> ![](https://i.imgur.com/Hzvd1c3.png)
>
> The first graph corresponds to the polarization where the electric field vector is perpendicular to the plane of incidence and the other with its electric field vector parallel to the plane of incidence.

### Annotation 8 (page 3)

Beside the passage: “clothing fabric. Further work, both experimental and theoretical, is surely needed on this fascinating everyday phenomenon.”

**Annotation** by Luís Batalha (2018-10-29):

> The question of why a soaked fabric is more transparent is slightly different.
>
> The fabric is made of many fibers with air in between and this structure causes light to bounce around many times inside, making it hard for light to get through.
>
> When we soak the fabric, we are replacing the air with water, which has a closer index of refraction to the fibers, making the inside reflections less important and more light goes through.

## Full text

Extracted from the PDF. [Annotation N] marks where each annotation above sits.

#### Page 1

Why some things are darker when wet

John Lekner and Michael C. Dorf

Angstrom has proposed that rough absorbing materials [Annotation 1] are darker when wet because their diffuse reflection makes possible total internal reflection in the water film covering them, increasing the likelihood of the absorption of light by the surface. His model is extended here in two ways: the probability of internal reflection is calculated more accurately, and the effect on absorption of the decrease of the relative refractive index (liquid to material instead of air to material) is estimated. Both extensions decrease the albedo of the wetted surface, bringing the model into good agreement with experiment.

1. Introduction

From early childhood we learn to distinguish wet from dry, not just by touch, but also by sight. Most objects, especially those with rough and absorbing surfaces, are darker when wet: they reflect less light. In a model for this phenomenon proposed by Angstrom,1 the surface roughness leads to diffuse reflection, and thus to total internal reflection at the liquid-air interface of the thin film of liquid covering the surface. This reflection gives an increased probability of the absorption of light by the surface, and thus a darker surface.

Angstrom's model is extended in the following two sections. The results are then compared with experiment. In the final section we also compare Angstrom's approach to that of Bohren 2 and Twomey et al. 3

(TBM), in which the darkening is taken to be due to the increase in the forward scattering on wetting.

II. Diffuse Scattering and Internal Reflection An optically smooth surface reflects specularly, whether or not the surface is covered by a film of liquid. A ray of light reflecting at a smooth surface will not be totally [Annotation 3] reflected at the liquid-air interface. If the surface is rough, its diffuse reflection sends some light out obliquely enough to be totally internally reflected at the liquid-air interface. This increases the chance of absorption at the surface.

The authors are with Victoria University of Wellington, Physics Department, Wellington, New Zealand. Received 6 June 1987. 0003-6935/88/071278-03$02.00/0. © 1988 Optical Society of America.

Consider a rough absorbing surface, such as a blackboard. We treat the surface as an array of randomly oriented facets, each of which reflects specularly. When wet, a thin liquid layer covers the surface. Light incident on this layer has probability 1 - R 1 of reaching the surface, where R 1 is the reflectance at the air-liquid interface. Some fraction of the transmitted light is absorbed by the surface. Call this fraction a. Of the light which is reflected by the surface, let p be the fraction which is then reflected back at the liquid-air interface, so that it is once again incident on the surface. The process is illustrated in Fig. 1. Continuing the process illustrated ad infinitum, the total probability of absorption by the rough surface is

A= (1-Rl)[a + a(l-a)p + a(l-a) 2 p 2 +...]

(1 [Annotation 2] - R)a 1 - p(l-a) (1)

Angstrom evaluates p as follows: all the light with an angle of incidence greater than the critical angle , (= arcsinl/n) onto the liquid-air interface will be totally reflected. Thus p may be estimated as the fraction of diffusely reflected light which lies outside the cone generated by rays whose reflection angle is 60. For [Annotation 4] a Lambertian surface, the intensity reflected at angle is proportional to cos6. Light emerging at angles [Annotation 5] to + d subtends a solid angle 27r sinOdO. Thus

i/2 27r dO sinG cosO

P p /2 = Cos [Annotation 6] 2 OC = 1- 1/n . 2,r I dO sinG cosO

Equations (1) and (2) are together equivalent to the last equation in Angstrom's Sec. 4(ii), except that he omits the 1 - R factor.

1278 APPLIED OPTICS / Vol. 27, No. 7 / 1 April 1988

(2)

#### Page 2

Fig. 1. Liquid layer over a rough surface. The coefficients represent the fraction of the incident light intensity which is transmitted along each path.

Equation (2) underestimates p: the reflectivities for both polarizations are generally small but not zero for 0 < 0,; see, for example, Ref. 4, Sec. 1.5, or Ref. 5, Figs. 1-3 and 1-4. For a surface which reflects diffuse- [Annotation 7] ly according to Lambert's law, the probability reflection at the liquid-air interface is J r/2 | dO sinG cosOR(x,1/nj) /2 X dO sinG cosO

= J dxR(x, /ni),

where x = sin 2 o and R(x,n) is the reflectance at an interface between media 1 and 2, with n = n 2 /ni. Here n = 1/ni < 1 and R is unity for x > n 2 so n2 p 1 - n 2 + dxR(x,n). (4)

For both polarizations, the reflectances have the property that

R(x,n) = R(x/n 2 , i/n) (5)

(this was noted by Stern 6 for the transmittances). On changing to the variable y = x/n 2 , the integral in Eq. (4) may be written as

n 2 j dyR(n 2 yn) = n 2 J dyR(y,i/n) = n 2 (/n). (6)

The last equality expressed the integral in terms of the average reflectance of an isotropically illuminated surface,

r-l2 J dO sinO cosOR(x,n) I R(n) = °= I dxR(x,n) (7, j dO sinG cosO °

(in correspondence with the average transmittance defined by Stern 6 ). For unpolarized light, Stern's formulas (9a) and (9b) lead to - 3n 2 +2n+i 2n 3 (n 2 + 2n-1) n 2 (n2 + 1) 3(n + 1)2 (n 2 + i) 2 (n 2 - 1) (n 2 - 1)2

n2- 1)2 n(n + 1) (n > 1). (8 (n 2 + i)3 l o g

From Eqs. (4) and (6) we have the result

p = 1- I [1- R(nj)]. n

(9)

For water, this formula gives a p larger by -9% than the Angstrom estimate (0.475 instead of 0.437).

Ill. Probability of Absorption by a Wetted Surface The parameter a in Eq. (1) is the fraction of the light incident on the surface which is absorbed. (The above refers to a single interaction: the total probability of absorption, allowing for reflections at the liquid-air interface, is A.) Angstrom takes a to have the same value for the wet as for the dry surface. We expect a, (the value when wet), to be greater than ad (the value when dry), since the absorbing medium will normally have the real part of its refractive index greater than unity. Since reflection is caused by wave vector mismatch, and since wave vectors are determined by refractive indices, covering the surface with a layer of liquid (with nj > 1) results in less reflection. The value of ad is in principle determined by the complex refractive index nr + ins of the material, and by the roughness of the surface (which influences the average angle of incidence on its randomly oriented facets, and the probability of multiple interactions, as in a crevice). The value of a, is in addition a function of nj, the refractive index of the liquid film covering it. For the purpose of comparing the albedos 1 - ad and 1 - A of the dry and wetted surfaces, we estimate a, in terms of ad, nl, and nr as follows. For small absorption (ni << nr), ad 1 - R(n) where R(n) is the average reflectance of an isotropically illuminated surface, defined in Eq. (7). The assumption made here is that the angle of incidence on facets of the rough surface (for, say, normal illumination) has the same distribution as would be obtained for a plane surface illuminated isotropically. Similarly, a, - 1 - (nr/nl) when ni << n,. Thus when the absorption is small,

a, ad[1 - R(nr/nl)]/[l - R(n_)] = a). (10)

When the absorption is large, on the other hand, and ad 1, we expect aw ad a(". We will use an interpolation formula which incorporates these limiting forms by giving a(°) and a(') the weights 1 - ad and ad:

aw (1 - ad)a(°) + ada(l),

or (11) ad ~ ~~ 1-R(n,1n 1 ) (I-d - + adad (1 ad) -R(nl)

Because R(n) is a monotonically increasing function of n (for n > 1), R(nr/nli) < R(nr) for ni < nr, and so aw > ad, with the greatest percentage increase occurring at low absorption. For small ad and ni = 4/3 (water), the ratio of aw/ad takes the values 1.07, 1.08, and 1.10 for nr = 1.5, 2, and 2.5 (most minerals have refractive indices within this range). When we put a,, as defined by (11), for a in Eq. (1), we can find the ratio of the total absorption by a wetted surface to the absorption by the dry surface, A/ad. This provides one measure of how much darker a given surface becomes when wetted.

1 April 1988 / Vol. 27, No. 7 / APPLIED OPTICS 1279

#### Page 3

The ratio A/ad is plotted as a function of nj (for two values of ad) in Fig. 2. We note from Fig. 2 that the darkening effect is strongest when the absorption is weak. When the absorption is strong the contribution of internal reflections is less important, since a larger fraction of the light is absorbed on first contact with the surface. The interpolation formula (11) gives a ad. We expect this to be true for all cases where the refractive index of the liquid film lies between that of air and the real part of the index of the surface. Now from Eq. (1) it follows that

dA (1 - p)(1 - R 1 ) > da [1 -p(i - a)]2 (12)

for all physical values of a, p, and R,. Thus taking an = ad provides a lower bound for A. This is shown by the dashed curves in Fig. 2, which neglect the increase in absorption due to the change in relative refractive index at the surface on wetting. As such, they isolate the effect of internal reflection, which is seen to be dominant for a weakly absorbing rough surface.

IV. Comparison with Experimental Data and Discussion Albedo is defined and measured as the ratio of the light intensity incident on a surface to that diffusely reflected back by the surface. For example, Angstrom measured albedo by taking the reading (a) of his upward facing pyranometer, then the reading () of the same pyranometer facing the ground. The albedo is / a. His Table III lists albedo comparisons for dry and wet sand (0.182 and 0.091) and for dry and wet black mold (0.141 and 0.084). The variability is indicated by the last value being an average of 0.091, 0.081, and 0.081. These are plotted as circles in Fig. 3, on which we compare theoretical albedos (1 - ad for the dry surface, 1 - A for the wet). Other data available are those listed by Sellers, 7 Table IV, for dry and wet sand dunes, and dry and wet savanna (grassland). These appear, respectively, as the large and small crosses in Fig. 3, the size of the crosses indicating the range of values observed. The agreement of Angstrom's theory (as extended here) with the data is similar to that obtained by Twomey et al. 3 (TBM), as shown in their Fig. 5. Angstrom's original theory, incorporating only the possibility of internal reflection (and that approximately) is seen to show the same trend, but agreement with the data is not quite so good. [However, the variation in the darkening of sand with the refractive index of the wetting liquid is not so strong in the Angstrom model as in the experimental data of TBM (Ref. 3, Fig. 7).] The fact that the two theories, based on such different models, both agree with the (very limited) wet and dry albedo data is surprising. Angstrom's theory would seem to apply best to rough solid surfaces, such as blackboards, asphalt, or concrete. The TBM theory, based on the idea of enhanced forward scattering when the interstitial space in the medium is filled with liquid, would seem to apply best to finely divided media, such as sand. An intermediate case is that of

3

Aa ad

1

0

1 2

Fig. 2. Ratio of average absorption by a wet surface to that of the dry surface, for varying refractive index of the wetting liquid. For the solid curves a,,, is given by (11), with nr = 2. The dashed curves are drawn for a,, = ad (Angstrom). Inall cases we have setRl = (nl -1)2/ (ni + 1) 2 (normal illumination).

wet albedo

0

0 dry albedo 1 Fig. 3. Wet albedo as a function of dry albedo, for a layer of water over a rough surface. Normal illumination is assumed. Dry albedo is given by 1 - ad and wet albedo by 1 - A, the latter calculated using nr = 2. A is given by Eq. (i); the solid line has a = a, as given by (ii), and p by Eq. (9). The dashed line has a = ad and p given by (2) (Angstrom's approximations). The experimental data are as described in the text.

clothing fabric. Further work, both experimental and [Annotation 8] theoretical, is surely needed on this fascinating everyday phenomenon.

We wish to thank Malcolm Wright and Joe Trodahl for stimulating discussions and Craig F. Bohren for critical comments on the original version of this paper.

References 1. A. Angstrom, "The Albedo of Various Surfaces of Ground," Geogr. Ann. 7, 323 (1925). 2. C. F. Bohren, "Multiple Scattering at the Beach," Weatherwise (Aug. 1983). 3. S. A. Twomey, C. F. Bohren, and J. L. Mergenthaler, "Reflectance and Albedo Differences Between Wet and Dry Surfaces," Appl. Opt. 25, 431 (1986). 4. M. Born and E. Wolf, Principles of Optics (Pergamon, Oxford, 1965). 5. J. Lekner, Theory of Reflection of Electromagnetic and Particle Waves (Martinus Nijhoff, Dordrecht, 1987). 6. F. Stern, "Transmission of Isotropic Radiation Across an Interface Between Two Dielectrics," Appl. Opt. 3, i (1964). 7. W. D. Sellers, Physical Climatology (U. Chicago Press, 1965).

1280 APPLIED OPTICS / Vol. 27, No. 7 / 1 April 1988

---

Source: [Fermat's Library](https://fermatslibrary.com/) — https://fermatslibrary.com/s/why-some-things-are-darker-when-wet. More annotated papers: https://fermatslibrary.com/llms.txt
